CONTROL OF THREE PHASE CONVERTERS AS SOURCE FOR MICROGRID
Mechatronics, PhD Thesis, 2017
Prof. Dr. Asif ŞABANOVİÇ (Thesis Advisor), Prof. Mustafa Ünel, Asst. Prof. Ahmet Teoman Naskalı, Asst. Prof. Eray Abdurrahman Baran, Asst. Prof. Ayesha Asloob Qureshi
Date & Time: July 6, 2017 - 11:30 am
Place: FENS 2008
Keywords: Grid Converters, Grid-Connected Source-Converter (GCSC), Renewable Energy Sources (RES), Micro-grid (MG), Smart Grid (SG), Symmetrical Components, Disturbance Observer (DOB), Equivalent Control, Unbalanced Grid, Voltage Sag/Swell
The demand for electricity is increasing day by day in the mordern world and the fossil fuels are ending rapidly. Alternatively, the renewable energy resources (RESs) especially wind and solar are being widely adopted as reliable electric sources. However, both the wind and the solar energies are highly unpredictable due to weather conditions and the both can not maintain their intensity constantly as required by the grid. So, voltage variations and power fluctuations are inevitable in the grid connected systems based on famous RESs. These fluctuations are very hazardous for the electrical loads and other connected components. In extreme, these can even make the whole distributed energy system to collapse. Also these reasons do not allow the integration of the RESs and thus, causing hurdles in realizing the smart grid (SG), which is the next generation of conventional grid due to its bidirectional power flow capabilities. Due to the same reasons, the grid code requirements demand the distributed generators (DGs) or Source Converters to stay connected and support the grid under all kind of its varying conditions. So, the grid-connected source-converters (GCSC) are required to be made dynamically controllable three-phase sources of energy compatible with varying conditions of the loads or distributed grid-connected network.
Since, the power electronic converters are the main components which play the role of interconnection between the source(s) and the microgrid (MG) or the grid so, the control of these power converters is very crucial to obtain reliable operation of the DG (or GCSC) and the overall distributed power system (DPS). The control of the converters is actually dependent on the control of its switching matrix. In general, the DC to AC converter has three independent control inputs and in majority of the control approaches, the goal is stated as control of the current (specified in orthogonal frame of references) or the power (active and reactive) thus having two dimensional control error. This leads to underutilization of the capabilities of the switching converter. So, in this thesis, one more control requirement along the current or power control has been investigated and defined to achieve a better control of three-phase power converter in unbalanced grid voltage conditions. The resultant controller of the converter has appeared to be a generalized solution, capable of generating/supplying the balanced and unbalanced voltages and/or compensating for the balanced and unbalanced current flow depending on the power grid status while maintaining the desired power generations too.
The role of zero-sequence voltage and zero-squence current has been developed since one of these (or both) does exist depending on the topology and normal or abnormal status of the grid. It defines the third control requirement in the developed equivalent control method, which by its nature does not have to rely on accurate knowledge of the grid or connected components’ parameters.
Finally, an improved grid current control strategy for grid connected voltage source inverters under unbalanced grid voltage conditions is also developed, which which satisfies the positive sequence power requirements and eliminates selective negative sequence power components, causing double frequency oscillations in power. The symmetrical components are calculated by using second order generalized integrator (SOGI) based observer, which accurately estimates the symmetrical components even when the voltage is disturbed. This proposed method also does not require any grid parameter to be known accurately since it estimates nonlinear terms with a first order low pass filter disturbance observer (DOB).